Integration of differential wavefront sampling with merit function regression for efficient alignment of three-mirror anastigmat optical system

摘要

We first studied the characteristics of alignment performances of two computer-aided alignment algorithms i.e. merit function regression (MFR) and differential wavefront sampling (DWS). The initial study shows i) that, utilizing damped least square algorithm, MFR offers accurate alignment estimation to the optical systems with non-linear wavefront sensitivity to changes in alignment parameters, but at the expense of neglecting the coupling effects among multiple optical components, and ii) that DWS can estimate the alignment state while taking the inter-element coupling effects into consideration, but at the expense of increased sensitivity to measurement error associated with experiment apparatus. Following the aforementioned study, we report a new improved alignment computation technique benefitted from modified MFR computation incorporating the concept of standard DWS method. The optical system used in this study is a three-mirror anastignmat (TMA) based optical design for the next generation geostationary ocean color instrument (GOCI-II). Using an aspheric primary mirror of 210 mm in diameter, the F/7.3 TMA design offers good imaging performance such as 80% in 4 um in GEE, MTF of 0.65 at 65.02 in Nyquist frequency. The optical system is designed to be packaged into a compact dimension of 0.25m * 0.55m x 1.050m. The trial simulation runs demonstrate that this integrated alignment method show much better alignment estimation accuracies than those of standard MFR and DWS methods, especially when in presence of measurement errors. The underlying concept, computational details and trial simulation results are presented together with implications to potential applications.